This dataset was scraped from nextspaceflight.com and includes all the space missions since the beginning of Space Race between the USA and the Soviet Union in 1957!

Installing Package with Country Codes

%pip install iso3166

Requirement already satisfied: iso3166 in c:\users\dacs\anaconda3\lib\site-packages (2.0.2)
Note: you may need to restart the kernel to use updated packages.

Upgrading Plotly

%pip install --upgrade plotly

Requirement already satisfied: plotly in c:\users\dacs\anaconda3\lib\site-packages (5.5.0)
Requirement already satisfied: tenacity>=6.2.0 in c:\users\dacs\anaconda3\lib\site-packages (from plotly) (8.0.1)
Requirement already satisfied: six in c:\users\dacs\anaconda3\lib\site-packages (from plotly) (1.16.0)
Note: you may need to restart the kernel to use updated packages.

Importing Libraries

import numpy as np
import pandas as pd
import plotly.express as px
import matplotlib.pyplot as plt
import seaborn as sns
from iso3166 import countries
from datetime import datetime, timedelta

Notebook Presentation Format

pd.options.display.float_format = '{:,.2f}'.format

Loading the Dataset

df_data = pd.read_csv('mission_launches.csv')

Preliminary Data Exploration

df_data.shape   # The shape is 4324 Rows and 9 Columns.
df_data.columns

Index(['Unnamed: 0', 'Unnamed: 0.1', 'Organisation', 'Location', 'Date',
       'Detail', 'Rocket_Status', 'Price', 'Mission_Status'],
      dtype='object')

Data Cleaning

df_data.isna().values.any() 

True

df_data.isna().sum()

Unnamed: 0           0
Unnamed: 0.1         0
Organisation         0
Location             0
Date                 0
Detail               0
Rocket_Status        0
Price             3360
Mission_Status       0
dtype: int64

As we can see, we cannot drop NaN values here, since the price is missing in about 75% of the rows, so we will deal with that once we start exploring the price column.

df_data.duplicated().values.any()

False

df_missions = df_data.drop(columns=['Unnamed: 0', 'Unnamed: 0.1'])

df_missions.head()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status
0	SpaceX	LC-39A, Kennedy Space Center, Florida, USA	Fri Aug 07, 2020 05:12 UTC	Falcon 9 Block 5 | Starlink V1 L9 & BlackSky	StatusActive	50.0	Success
1	CASC	Site 9401 (SLS-2), Jiuquan Satellite Launch Ce...	Thu Aug 06, 2020 04:01 UTC	Long March 2D | Gaofen-9 04 & Q-SAT	StatusActive	29.75	Success
2	SpaceX	Pad A, Boca Chica, Texas, USA	Tue Aug 04, 2020 23:57 UTC	Starship Prototype | 150 Meter Hop	StatusActive	NaN	Success
3	Roscosmos	Site 200/39, Baikonur Cosmodrome, Kazakhstan	Thu Jul 30, 2020 21:25 UTC	Proton-M/Briz-M | Ekspress-80 & Ekspress-103	StatusActive	65.0	Success
4	ULA	SLC-41, Cape Canaveral AFS, Florida, USA	Thu Jul 30, 2020 11:50 UTC	Atlas V 541 | Perseverance	StatusActive	145.0	Success

Descriptive Statistics

df_missions.describe()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status
count	4324	4324	4324	4324	4324	964	4324
unique	56	137	4319	4278	2	56	4
top	RVSN USSR	Site 31/6, Baikonur Cosmodrome, Kazakhstan	Wed Nov 05, 2008 00:15 UTC	Cosmos-3MRB (65MRB) | BOR-5 Shuttle	StatusRetired	450.0	Success
freq	1777	235	2	6	3534	136	3879

df_organisations = df_missions.groupby('Mission_Status', as_index=False).agg({'Organisation': pd.Series.count})
df_organisations.sort_values('Organisation', ascending=False, inplace=True)
df_organisations

	Mission_Status	Organisation
3	Success	3879
0	Failure	339
1	Partial Failure	102
2	Prelaunch Failure	4

Number of Launches per Company

df_nr_launches = df_missions.groupby('Organisation', as_index=False).agg({'Mission_Status': pd.Series.count})
df_nr_launches.head()

	Organisation	Mission_Status
0	AEB	3
1	AMBA	8
2	ASI	9
3	Arianespace	279
4	Arm??e de l'Air	4

fig = px.bar(x=df_nr_launches.Organisation,
             y=df_nr_launches.Mission_Status,
             color=df_nr_launches.Mission_Status,
             title='Number of Launches per Company')

fig.update_layout(xaxis_title='Organisation', 
                    coloraxis_showscale=False,
                    yaxis_title='Number of Launches')
fig.show()

Number of Active versus Retired Rockets

df_rockets = df_missions.groupby('Rocket_Status', as_index=False).count()

df_rockets.value_counts()

Rocket_Status  Organisation  Location  Date  Detail  Price  Mission_Status
StatusActive   790           790       790   790     586    790               1
StatusRetired  3534          3534      3534  3534    378    3534              1
dtype: int64

Distribution of Mission Status

Checking how many missions were sucessfull and how many failed.

df_mission_status = df_missions.groupby('Mission_Status', as_index=False).count()

df_mission_status.value_counts()

Mission_Status     Organisation  Location  Date  Detail  Rocket_Status  Price
Failure            339           339       339   339     339            36       1
Partial Failure    102           102       102   102     102            17       1
Prelaunch Failure  4             4         4     4       4              1        1
Success            3879          3879      3879  3879    3879           910      1
dtype: int64

How Expensive are the Launches?

The price column is given in USD millions (we need to take care of NaN's here).

df_price_of_launches = df_data.dropna()
df_price_of_launches.Price.isna().values.any()

False

df_price_of_launches_sorted = df_price_of_launches.sort_values('Price', ascending=True)

hist = px.histogram(df_price_of_launches_sorted, x='Price',
                    nbins=40,
                    opacity=0.6)

hist.update_layout(xaxis_title='Price of Launch',
                   yaxis_title='Count')

hist.show()

Number of Launches by Country

Had to work the country names so I could get a consistent naming structure to chart it. (Used the iso31666 package)

df_country_launches = df_data.groupby(['Location'], as_index=False).agg({'Mission_Status': pd.Series.count})
df_country_launches['Location'] = df_country_launches.Location.str.split(',').str[-1]
df_country_launches.head()

	Location	Mission_Status
0	USA	12
1	France	4
2	USA	1
3	USA	6
4	France	15

df_missions['launch_country'] = df_missions["Location"].str.split(", ").str[-1]

df_missions['launch_country'].replace({'Gran Canaria': 'USA', 'Yellow Sea': 'China', 
                                  'Pacific Missile Range Facility': 'USA', 'Barents Sea': 'Russian Federation', 
                                  'Russia': 'Russian Federation', 'Pacific Ocean': 'USA',
                                  'Marshall Islands': 'USA', 'Iran': 'Iran, Islamic Republic of', 
                                  'North Korea': "Korea, Democratic People's Republic of", 'South Korea': "Korea, Republic of",
                                  'Shahrud Missile Test Site': "Iran, Islamic Republic of", 'New Mexico': "USA" }, 
                                  inplace=True)

df_missions['launch_country_code'] = df_missions['launch_country'].apply(lambda x: (countries.get(x).alpha3))

df_launches = df_missions.groupby(['launch_country', 'launch_country_code'], as_index=False).agg({'Mission_Status': pd.Series.count})
df_launches.rename(columns={'Mission_Status': 'Total_launches'}, inplace=True)

fig = px.choropleth(data_frame=df_launches,
              locations='launch_country_code',
              color='Total_launches',
              color_continuous_scale='matter')

fig.update_layout(coloraxis_showscale=True,)

fig.show()

Number of Failures by Country

df_country_fails = df_missions.where(df_missions.Mission_Status != 'Success')
df_country_fails = df_country_fails.groupby(['launch_country', 'launch_country_code'], as_index=False).agg({'Mission_Status': pd.Series.count})
df_country_fails.rename(columns={'Mission_Status': 'Total_launches'}, inplace=True)

df_country_fails.head()

	launch_country	launch_country_code	Total_launches
0	Australia	AUS	3
1	Brazil	BRA	3
2	China	CHN	25
3	France	FRA	18
4	India	IND	13

fig2 = px.choropleth(df_country_fails, locations='launch_country_code',
                    hover_name="Total_launches", # column to add to hover information
                    color_continuous_scale=px.colors.sequential.matter)

fig2.update_layout(coloraxis_showscale=True,)

fig2.show()

Countries, organisations, and mission status.

country_org_status = df_missions.groupby(['launch_country_code', 'Mission_Status', 'Organisation'], as_index=False).agg({'Mission_Status': pd.Series.count})
country_org_status

	launch_country_code	Organisation	Mission_Status
0	AUS	CECLES	2
1	AUS	RAE	1
2	AUS	AMBA	1
3	AUS	CECLES	1
4	AUS	RAE	1
...	...	...	...
127	USA	Sea Launch	33
128	USA	SpaceX	94
129	USA	ULA	139
130	USA	US Air Force	129
131	USA	US Navy	2

132 rows × 3 columns

burst = px.sunburst(country_org_status, 
                    path=['launch_country_code', 'Organisation', 'Mission_Status'], 
                    values='Mission_Status',
                    title='Countries, organisations, and mission status.',
                   )
 
burst.update_layout(xaxis_title='Number of Missions', 
                    yaxis_title='Organisation',
                    coloraxis_showscale=False)
 
burst.show()

Total Amount of Money Spent on Space Missions

Without NaN values

df_missions.Price = pd.to_numeric(df_missions.Price, errors = 'coerce') #'coerce' means that invalid parsing will be set as NaN.
df_missions.Price = df_missions.Price.fillna(0)
df_missions.Price.sum()

123175.68000000002

Amount of Money Spent by Organisation

df_missions.Price.describe()

count   4,324.00
mean       28.49
std        85.93
min         0.00
25%         0.00
50%         0.00
75%         0.00
max       450.00
Name: Price, dtype: float64

org_price = df_missions.groupby(['Price', 'Organisation'], as_index=False).agg({'Price': pd.Series.sum})
org_price.sort_values('Price', ascending=False, inplace=True)
org_price

	Organisation	Price
122	NASA	61,200.00
119	Arianespace	15,000.00
101	ULA	3,706.00
121	ULA	3,500.00
98	MHI	2,520.00
...	...	...
45	UT	0.00
44	US Navy	0.00
43	US Air Force	0.00
42	ULA	0.00
0	AEB	0.00

123 rows × 2 columns

fig3 = px.bar(x=org_price.Organisation,
             y=org_price.Price,
             title='Money Spent by Organisation')

fig3.update_layout(xaxis_title='Organisation', 
                    coloraxis_showscale=False,
                    yaxis_title='Money Spent')
fig3.show()

Number of Launches per Year

df_missions.Date = pd.to_datetime(df_missions.Date, utc=True)

df_missions_launch = df_missions.groupby(df_missions.Date.dt.year).count()

plt.figure(figsize=(16,8), dpi=200)

plt.title('Number of Launches per Year', fontsize=18)

plt.scatter(x=df_missions_launch.Date.index,
            y=df_missions_launch.Date.values,
            color='dodgerblue',
            alpha=0.7,
            s=100)
plt.show()

Number of Launches Month-on-Month until the Present

df_missions['year'] = pd.DatetimeIndex(df_missions['Date']).year
df_missions['month'] = pd.DatetimeIndex(df_missions['Date']).month
df_missions.head()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status	launch_country	launch_country_code	year	month
0	SpaceX	LC-39A, Kennedy Space Center, Florida, USA	2020-08-07 05:12:00+00:00	Falcon 9 Block 5 | Starlink V1 L9 & BlackSky	StatusActive	50.00	Success	USA	USA	2020	8
1	CASC	Site 9401 (SLS-2), Jiuquan Satellite Launch Ce...	2020-08-06 04:01:00+00:00	Long March 2D | Gaofen-9 04 & Q-SAT	StatusActive	29.75	Success	China	CHN	2020	8
2	SpaceX	Pad A, Boca Chica, Texas, USA	2020-08-04 23:57:00+00:00	Starship Prototype | 150 Meter Hop	StatusActive	0.00	Success	USA	USA	2020	8
3	Roscosmos	Site 200/39, Baikonur Cosmodrome, Kazakhstan	2020-07-30 21:25:00+00:00	Proton-M/Briz-M | Ekspress-80 & Ekspress-103	StatusActive	65.00	Success	Kazakhstan	KAZ	2020	7
4	ULA	SLC-41, Cape Canaveral AFS, Florida, USA	2020-07-30 11:50:00+00:00	Atlas V 541 | Perseverance	StatusActive	145.00	Success	USA	USA	2020	7

cols=["year","month"]
df_missions['year-month'] = df_missions[cols].apply(lambda x: '-'.join(x.values.astype(str)), axis="columns")
df_missions.head()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status	launch_country	launch_country_code	year	month	year-month
0	SpaceX	LC-39A, Kennedy Space Center, Florida, USA	2020-08-07 05:12:00+00:00	Falcon 9 Block 5 | Starlink V1 L9 & BlackSky	StatusActive	50.00	Success	USA	USA	2020	8	2020-8
1	CASC	Site 9401 (SLS-2), Jiuquan Satellite Launch Ce...	2020-08-06 04:01:00+00:00	Long March 2D | Gaofen-9 04 & Q-SAT	StatusActive	29.75	Success	China	CHN	2020	8	2020-8
2	SpaceX	Pad A, Boca Chica, Texas, USA	2020-08-04 23:57:00+00:00	Starship Prototype | 150 Meter Hop	StatusActive	0.00	Success	USA	USA	2020	8	2020-8
3	Roscosmos	Site 200/39, Baikonur Cosmodrome, Kazakhstan	2020-07-30 21:25:00+00:00	Proton-M/Briz-M | Ekspress-80 & Ekspress-103	StatusActive	65.00	Success	Kazakhstan	KAZ	2020	7	2020-7
4	ULA	SLC-41, Cape Canaveral AFS, Florida, USA	2020-07-30 11:50:00+00:00	Atlas V 541 | Perseverance	StatusActive	145.00	Success	USA	USA	2020	7	2020-7

df_nr_launches_month = df_missions.groupby('year-month', as_index=False).agg({'Mission_Status': pd.Series.count})
df_nr_launches_month

	year-month	Mission_Status
0	1957-10	1
1	1957-11	1
2	1957-12	1
3	1958-10	3
4	1958-11	1
...	...	...
742	2020-4	5
743	2020-5	9
744	2020-6	7
745	2020-7	14
746	2020-8	3

747 rows × 2 columns

fig3 = px.bar(x=df_nr_launches_month['year-month'],
             y=df_nr_launches_month.Mission_Status,
             title='Number of Launches by month')

fig3.update_layout(xaxis_title='Number of Launches', 
                    coloraxis_showscale=False,
                    yaxis_title='Month')
fig3.show()

Which months are most popular and least popular for launches?

df_nr_launches_month.sort_values(by='Mission_Status', ascending=False)

	year-month	Mission_Status
156	1971-12	18
212	1975-9	16
236	1977-9	16
730	2019-12	16
123	1968-4	16
...	...	...
550	2004-10	1
272	1980-9	1
271	1980-8	1
62	1963-3	1
0	1957-10	1

747 rows × 2 columns

df_missions['month'].value_counts()

12    450
6     402
4     383
10    381
8     373
9     365
3     353
7     351
2     336
11    336
5     326
1     268
Name: month, dtype: int64

How has the Launch Price varied Over Time?

launch_price = df_missions.groupby(['year'], as_index=False).agg({'Price': pd.Series.sum})

fig2 = px.scatter(x=launch_price.year,
             y=launch_price.Price,
             title='Launch Price Over Time')

fig2.update_layout(xaxis_title='Categories', 
                    coloraxis_showscale=False,
                    yaxis_title='Launch Price')
fig2.show()

Number of Launches over Time by the Top 10 Organisations.

nr_launches_org = df_missions.groupby(['year', 'Organisation'], as_index=False).agg({'Mission_Status': pd.Series.count})
nr_launches_org

	year	Organisation	Mission_Status
0	1957	RVSN USSR	2
1	1957	US Navy	1
2	1958	AMBA	7
3	1958	NASA	2
4	1958	RVSN USSR	5
...	...	...	...
658	2020	Roscosmos	4
659	2020	SpaceX	14
660	2020	ULA	4
661	2020	VKS RF	3
662	2020	Virgin Orbit	1

663 rows × 3 columns

fig2 = px.scatter(x=nr_launches_org.year,
             y=nr_launches_org.Organisation,
             title='Number of Launches over Time by the Top 10 Organisations',
             color=nr_launches_org.Mission_Status)

fig2.update_layout(xaxis_title='Year', 
                    coloraxis_showscale=False,
                    yaxis_title='Organisation')
fig2.show()

Cold War Space Race: USA vs USSR

The cold war lasted from the start of the dataset up until 1991.

cold_war = df_missions.where(df_missions.year < 1991)

cold_war = cold_war.loc[(cold_war['launch_country_code'] == 'USA') | (cold_war['launch_country_code'] == 'RUS') | (cold_war['launch_country_code'] == 'KAZ')]
cold_war.head()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status	launch_country	launch_country_code	year	month	year-month
1776	RVSN USSR	Site 32/2, Plesetsk Cosmodrome, Russia	1990-12-22 07:28:00+00:00	Tsyklon-3 | Cosmos 2114 to 2119	StatusRetired	0.00	Success	Russian Federation	RUS	1,990.00	12.00	1990-12
1777	RVSN USSR	Site 133/3, Plesetsk Cosmodrome, Russia	1990-12-10 07:54:00+00:00	Cosmos-3M (11K65M) | Cosmos 2112	StatusRetired	0.00	Success	Russian Federation	RUS	1,990.00	12.00	1990-12
1778	RVSN USSR	Site 90/20, Baikonur Cosmodrome, Kazakhstan	1990-12-04 00:48:00+00:00	Tsyklon-2 | Cosmos 2107	StatusRetired	0.00	Success	Kazakhstan	KAZ	1,990.00	12.00	1990-12
1779	NASA	LC-39B, Kennedy Space Center, Florida, USA	1990-12-02 06:49:00+00:00	Space Shuttle Columbia | STS-35	StatusRetired	450.00	Success	USA	USA	1,990.00	12.00	1990-12
1780	General Dynamics	SLC-3W, Vandenberg AFB, California, USA	1990-12-01 15:57:00+00:00	Atlas-E/F Star-37S-ISS | DMSP F-10	StatusRetired	0.00	Success	USA	USA	1,990.00	12.00	1990-12

Total number of launches of the USSR and the USA

Included Kazakhstan since at the time of the cold war it was part of USSR

cold_war_launches = cold_war.groupby(['launch_country_code'], as_index=False).agg({'Mission_Status': pd.Series.count})
cold_war_launches

	launch_country_code	Mission_Status
0	KAZ	578
1	RUS	1163
2	USA	644

fig = px.pie(cold_war_launches,
             labels=cold_war_launches.launch_country_code,
             values=cold_war_launches.Mission_Status,
             names=cold_war_launches.launch_country_code,
             title='Total Number of Launches from the Cold War',
             hole=0.3)

fig.update_traces(textposition='inside',
                  textfont_size=15,
                  textinfo='percent')

fig.show()

Total Number of Launches Year-On-Year by the Two Superpowers

cold_war_yearonyear = cold_war.groupby(['year', 'launch_country_code'], as_index=False).agg({'Mission_Status': pd.Series.count})
cold_war_yearonyear

	year	launch_country_code	Mission_Status
0	1,957.00	KAZ	2
1	1,957.00	USA	1
2	1,958.00	KAZ	5
3	1,958.00	USA	23
4	1,959.00	KAZ	4
...	...	...	...
93	1,989.00	RUS	22
94	1,989.00	USA	16
95	1,990.00	KAZ	7
96	1,990.00	RUS	30
97	1,990.00	USA	26

98 rows × 3 columns

fig3 = px.bar(x=cold_war_yearonyear['year'],
             y=cold_war_yearonyear.Mission_Status,
             color=cold_war_yearonyear.launch_country_code,
             barmode = 'group',
             title='Number of Launches by year')

fig3.update_layout(xaxis_title='Year', 
                    coloraxis_showscale=False,
                    yaxis_title='Number of Launches')
fig3.show()

Total Number of Mission Failures Year on Year.

mission_fails = cold_war.loc[(cold_war['Mission_Status'] == 'Failure')]
mission_fails.head()

	Organisation	Location	Date	Detail	Rocket_Status	Price	Mission_Status	launch_country	launch_country_code	year	month	year-month
1795	RVSN USSR	Site 45/2, Baikonur Cosmodrome, Kazakhstan	1990-10-04 04:27:00+00:00	Zenit-2 | Tselina-2 n†­8	StatusRetired	0.00	Failure	Kazakhstan	KAZ	1,990.00	10.00	1990-10
1837	Martin Marietta	SLC-40, Cape Canaveral AFS, Florida, USA	1990-03-14 11:52:00+00:00	Commercial Titan III | Intelsat 603	StatusRetired	136.60	Failure	USA	USA	1,990.00	3.00	1990-3
1885	RVSN USSR	Site 32/2, Plesetsk Cosmodrome, Russia	1989-06-09 10:10:00+00:00	Tsyklon-3 | Okean 2a	StatusRetired	0.00	Failure	Russian Federation	RUS	1,989.00	6.00	1989-6
2005	General Dynamics	SLC-36B, Cape Canaveral AFS, Florida, USA	1987-03-26 21:22:00+00:00	Atlas-G Centaur-D1AR | FLTSATCOM-6	StatusRetired	0.00	Failure	USA	USA	1,987.00	3.00	1987-3
2040	RVSN USSR	Site 32/2, Plesetsk Cosmodrome, Russia	1986-10-15 05:24:00+00:00	Tsyklon-3 | Cosmos 1786 to 1791	StatusRetired	0.00	Failure	Russian Federation	RUS	1,986.00	10.00	1986-10

hist2 = px.histogram(mission_fails, x='year',
                    nbins=40,
                    opacity=0.6)

hist2.update_layout(xaxis_title='Year',
                   yaxis_title='Number of Failed Launches')

hist2.show()

Percentage of Failures

cold_war['Mission_Status'].value_counts(normalize=True) * 100

Success             88.09
Failure              9.14
Partial Failure      2.73
Prelaunch Failure    0.04
Name: Mission_Status, dtype: float64

Lead country in terms of total number of launches for every year

launches_leaders_year = df_missions.groupby(['launch_country_code', 'year'], as_index=False).agg({'Mission_Status': pd.Series.count})
launches_leaders_year

	launch_country_code	year	Mission_Status
0	AUS	1967	1
1	AUS	1968	1
2	AUS	1969	1
3	AUS	1970	2
4	AUS	1971	1
...	...	...	...
407	USA	2016	27
408	USA	2017	30
409	USA	2018	34
410	USA	2019	27
411	USA	2020	21

412 rows × 3 columns

country_leaders_year = df_missions.groupby(['year', 'launch_country_code'])['Mission_Status'].count().reset_index().sort_values(['year', 'Mission_Status'], ascending=False)
country_leaders_year = pd.concat([group[1].head(1) for group in country_leaders_year.groupby(['year'])])
country_leaders_year.columns = ['year', 'launch_country_code', 'nr_launches']

fig5 = px.bar(
    country_leaders_year, 
    x="year", 
    y="nr_launches", 
    color='launch_country_code', 
    title='Leaders by launches for every year (countries)')

fig5.update_layout(xaxis_title='Year', 
                    coloraxis_showscale=False,
                    yaxis_title='Number of Launches')

fig5.show()

Organisation Doing the Most Number of Launches per year

org_leaders_year = df_missions.groupby(['year', 'Organisation'])['Mission_Status'].count().reset_index().sort_values(['year', 'Mission_Status'], ascending=False)
org_leaders_year = pd.concat([group[1].head(1) for group in org_leaders_year.groupby(['year'])])
org_leaders_year.columns = ['year', 'organisation', 'nr_launches']

fig6 = px.bar(
    org_leaders_year, 
    x="year", 
    y="nr_launches", 
    color='organisation', 
    title='Leaders by launches for every year (organisation)')

fig6.show()